Ollama Monitoring

Last updated - 07 January, 2026

LogicMonitor offers monitoring for Ollama language models using custom Groovy scripts and Ollama’s REST API. The integration benchmarks model performance by measuring token counts, model response latency, and load durations. This package enables proactive monitoring of AI infrastructure and ensures insight into model efficiency.

Requirements for Ollama Monitoring

• Install or update the Ollama LogicModules.
  For more information, see LogicModules in Package.
• Collector version 32.400 or higher.
• Access to the Ollama REST API.

Adding Ollama Resources into Monitoring

Ollama Resources are added to your environment manually. For more information, see Adding Resources.

Assigning Ollama Properties to Resources

The following custom properties must be set on your Ollama resource within LogicMonitor. For more information on setting properties, see Resource and Instance Properties.

Property	Description	Required?
system.categories	Must add Ollama to system.categories before monitoring to begin.	Yes
ollama.port	Specifies the port used to access the Ollama API for retrieving loaded models. Defaults to 11434.	Optional
ollama.models	Specifies the machine learning models to monitor or defaults to currently loaded models if not specified. Supports a comma separated list.	Yes
ollama.prompt	Specifies the prompt used to generate responses from models for benchmarking. Defaults to “how many letters are in LogicMonitor.”	Optional

Import LogicModules

From the LogicMonitor public repository, import all Ollama LogicModules, which are listed in the LogicModules in Package. If these LogicModules are already present, ensure you have the most recent versions. After the LogicModules are imported and the required property is in place, data collection and discovery begins automatically.

LogicModules in Package

LogicMonitor’s package for Ollama consists of the following LogicModules. For full coverage, make sure that all of these LogicModules are imported into your LogicMonitor platform.

Display Name	Typer	Description
Ollama Benchmark	DataSource	Benchmarking of language models currently loaded in Ollama.

When setting static datapoint thresholds on the various metrics tracked by this package’s DataSources, LogicMonitor follows the technology owner’s best practice KPI recommendations.